In a typical path of an optical communication system, an optical signal will be amplified by multiple optical amplifiers along the path. Optical amplifiers in communication systems are typically operated such that the gain of each amplifier is held constant using a control loop operating in accordance with a control algorithm. The control algorithms that have been used in various implementations follow several different approaches. For example, one approach is to operate at a constant gain unless the power change at the amplifier input exceeds some threshold value, at which point a rapid response algorithm is implemented. This helps overcome the lagging response of the amplifier to very rapid and large power changes that might occur, for example, due to a fiber break. For rapid adjustments, most algorithms respond with limited information based upon measurements of the total optical power input and/or output of the amplifier.
In wavelength-division multiplexing (WDM) optical transmission systems, the input power to optical amplifiers can change due to a wide variety of events such as the loss of WDM channels, faulty components, and upstream network power adjustments, among others. It is desirable in such systems for each WDM channel to maintain a particular target power and therefore the gain profile of the optical amplifiers is set to a value that depends on the number of WDM channels and their target powers. This is typically achieved by directly setting the amplifier to achieve the target output or by maintaining a fixed gain in the amplifier using a variety of different algorithms. At random times in the network operation, however, channels can be lost due to events such as, for example, an upstream fiber break. When channels are lost, the amplifiers must be adjusted down in power to the level appropriate for the new channel configuration. During this process, there can be error in the power setting and this gain error can potentially grow as the signals are transmitted through the network. This is particularly true when abrupt changes occur making it difficult for the system to rapidly adjust. During such rapid changes, it may not be possible to determine the actual power of each WDM channel at an amplifier input. Each amplifier must adjust its gain, therefore, based upon the total input and/or output power changes or other indicators and not directly using actual channel powers. This results in a greater likelihood of error.
Moreover, optical amplifiers may be distributed, with an amplifier's input and output located at two different geographic locations. When a failure occurs, the amplifier control mechanism must respond using information about only the output power. Because the control mechanism does not know the input power, it is impossible for it to maintain constant gain. Depending on the pre- and post-transient channel powers, it may be possible to reduce the total gain error, however, it is very difficult to have zero gain error for all possible channel configurations.
The rapid nature of transient events and the errors involved in responding thereto means that it is often impossible to communicate control information (e.g., channel powers) between different nodes when such events occur.